Quantum confinement in nanodimensional counterpart of silicon—leading to the discretization of both electron and phonon density of states—causes the shrinking of hot electron versus phonon interaction cross section. As a consequence, photogenerated hot-carriers get longer lifetime in silicon quantum structures compared to bulk silicon. In the present work, a machine learning prediction model utilizing XGBoost has been introduced in order to estimate the photoluminescence intensity decay based on significant factors, such as dimension of nanocrystals, excitation photon energy, photoluminescence emission energy, and photoluminescence intensity. The study emphasizes the benefits of employing machine learning, particularly XGBoost, in addressing complex nonlinear relationships while delivering interpretable outcomes. The results indicate that machine learning techniques can successfully model hot-carrier dynamics in silicon nanostructures, providing important insights on the immense potential of this material to be utilized in the light absorber layer of the hot-carrier solar cell.